INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lock

-0.87

kok

-0.77

nar

-0.77

riz

-0.71

pter

-0.70

ukong

-0.69

iverpool

-0.69

Kak

-0.68

chat

-0.68

engu

-0.67

POSITIVE LOGITS

 abdom

0.69

 Grizz

0.69

 Byrd

0.69

ADA

0.64

 Dull

0.63

 Bear

0.63

Ir

0.62

 Yellowstone

0.62

 ADHD

0.61

 Kinnikuman

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.